JP2006178757A - Distributed virtual environment management system and method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a distributed virtual environment (DVE) management system. <P>SOLUTION: A master server 21 receives simultaneous user redistribution requirement and a user number first tree from a first slave server 22, and receives a user number second tree from a second slave server 23. When deciding that a first tree first area first portion of the first slave server 22 can be controlled by the second slave server 22, a decision result wherein control of the first slave server 22 first area first portion is converted into the second slave server 23 is transmitted. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to distributed virtual environment (DVE) management, and more particularly, to a system and method of communication range control (Area of Internet Management) AOIM.

  A distributed virtual environment management system allows geographically dispersed users to exchange state information over a network such as the Internet, thereby exchanging information with each other and sharing a consistent view of the virtual environment (VE). is there. When a DVE system grows in terms of VE size and number of concurrent users, its scalability, i.e., more concurrent users can interact with a larger VE without compromising interactive performance. Support is an important issue. One approach to increase extensibility is the multiserver architecture, which has been adopted by many commercial multiplayer network games as well as some DVE systems. In order to manage the entire multi-server, dividing the VE into multi-regions and distributing the responsibilities can greatly reduce the work of individual servers. As a result, the system can support more concurrent users and larger VEs.

  The uneven distribution of users on DVE results in an imbalance of workload between servers, that is, some servers handle more congested areas than others and suffer from heavy workloads. As a result, users managed by heavy load servers suffer from low interactive performance due to long wait times for server state updates. A dynamic load distribution scheme is required to prevent degradation of interactive performance.

  An object of the present invention is to provide a distributed virtual environment DVE management system.

  The system includes a master server, a first slave server, and a second slave server. The master server receives the redistribution request and the first tree from the first slave server, and the second tree from the second slave server. The first tree is composed of a plurality of first nodes. Each first node consists of a first number of simultaneous users indicating how many users are in the first area or part of the first area. The second tree is composed of a plurality of second nodes. Each second node consists of a second number of simultaneous users indicating how many users are in the second area or part of the second area. The first part of the first area is close to the second area, and the number of second concurrent users in the second area plus the number of first concurrent users in the first area is the server load threshold or server If it is lower than the load threshold ratio, the master server determines that the first portion of the first region is controlled by the second slave server. The master server further transmits a result indicating that the portion of the first region is controlled by the second slave server, and switches control of the first portion of the first region between the first slave server and the second slave server. Let

  A DVE management method executed by the master server is provided. Computer code settings are also provided to implement DVE management methods.

  According to the present invention, the master server transmits a result indicating that the part of the first area is controlled by the second slave server, and controls the first part of the first area between the first slave server and the second slave server. Is converted.

  FIG. 1 is a diagram showing a specific example of a DVE system, which includes a master server 21, a backup server 26, and slave servers 22, 23, 24, 25, and is connected to a network (preferably the Internet or Operate on the intranet. Those skilled in the art recognize that the master server 21, the backup server 26, and the slave servers 22, 23, 24, 25 are connected to different types of network environments, and between different types of network environments, routers, gateways, Communicate with access points, base station systems, or various other types of transmission equipment. The four areas 31, 32, 33, and 34 of the VE are individually managed by the slave servers 22, 23, 24, and 25. It should be noted that the sizes of the regions are not equal, and their shapes are not limited to squares and rectangles, but various forms.

  FIG. 2 is a diagram illustrating a hardware environment applied to the master, backup, and slave servers 21 to 25. The processing unit 11, the memory 12, the storage device 13, the input device 14, the output device 15, and the communication device 16, Consists of. The processing unit 11 is connected to the memory 12, the storage device 13, the input device 14, the output device 15, and the communication device 16 through the bus 17 based on the von Neumann mechanism. A computer processor having one or more processing units 11 is composed of a single CPU, a microprocessor MPU, or a multi-processing unit, and is generally called a parallel processing environment. The memory 12 is preferably a RAM, but may be a ROM or a flash ROM. The memory 12 preferably records program modules executed by the processing unit 11 and executes the DVE management function. Generally, program modules include routines, programs, objects, components, scripts, web pages, or the like that perform detailed tasks or execute detailed abstract data types. In addition, those skilled in the art will recognize that certain embodiments are portable devices, multiprocessor-based, microprocessor-based, or programmable home appliances, network PCs, minicomputers, mainframe computers, and other computer system equipment. Understand what is practiced by composition. One implementation is implemented in a distributed computing environment where tasks are performed by remote processing devices that are linked to a communications network. In a distributed computing environment, program modules are located in both local and remote memory storage devices based on various remote access mechanisms such as DCOM, CORBA, web objects, web services, or other similar mechanisms. The storage device 13 is a hard drive, a magnetic device, an optical device, a portable device, or a nonvolatile memory device. Drivers and computer-readable media (as required) provide non-volatile computer-readable instructions, data structures, and program modules. The processing unit 11 controlled by the memory 12 and the program module received from the operator through the input device directs the DVE management function.

  FIG. 3 is a diagram illustrating a software mechanism according to a specific example of the DVE management system. Each of the slave servers 22 to 25 and the backup server 26 has local area management modules 221 to 261, and the master server 21 includes a VE management module 211. The management module of each area maintains the number of concurrent users in the local area allocated by the master server 21 by the quad tree, deletes the child node, adds a new child node, and based on the number, The management module 211 issues a re-segmentation request regarding the quadtree, and receives the redistribution result from the VE management module 211. The VE management module 211 performs a region retreat function, and based on the received quadtree, additional servers can be relocated to a region with a VE, if necessary, a retreat region processed by the slave servers 22-25, and It is determined whether it is necessary for processing of the backup server 26.

  Each node of the quad tree is divided into two dimensions and becomes four child nodes, corresponding to the whole / part of the local region. FIG. 4 is a diagram illustrating a typical quadtree of a local region including four portions R41a to R41d. The node N41 records the number of simultaneous users in the entire area R41, such as 31, and the nodes N41a, N41b, N41c, and N41d are simultaneously in the areas R41a, R41b, R41d, and R41c such as 5, 5, 5, and 16 Record the number of users. The four sub-regions of northwest NW, northeast NE, southeast SE, and southwest SW are arranged in order by the nodes N41a, N41b, N41c, and N41d. Graphic information of all nodes is recorded respectively.

  FIGS. 5a to 5d are flowcharts showing a method of DVE management, which is divided into two sections, and the left section showing the steps executed by the local area management modules 221, 231, 241, or 251 for easy understanding. Therefore, it is a right section showing the steps executed by the VE management module 211, separated by dotted lines.

  In step 411, a move message including information about the source and target areas is received, indicating that a user from the source area has moved to the target area. In step 421, it is determined whether the movement is coming from another server, and if so, step 453 is executed. Otherwise, step 423 is executed. The determination is accomplished by detecting if the source region is in a maintained quadtree. In step 423, a node corresponding to the received source region is placed. In step 425, the number of simultaneous users of the node is decreased by one. In step 431, it is determined that the placement node is the root node, and in that case, step 451 is executed. Otherwise, step 433 is performed. In step 433, the parent node is placed. In step 435, the concurrent users of the placement node are decreased one by one. In step 441, it is determined whether the child node of the placement node needs to be merged based on the updated number of concurrent users. For example, if the updated number of concurrent users is less than the region threshold, the child nodes of the placement node must be merged. Region thresholds are used to limit the number of concurrent users of a single leaf node. In step 443, the child node of the placement node is removed.

  In step 451, it is determined whether the move goes to another server, in which case the process ends. Otherwise, step 453 is performed. The determination is accomplished by detecting whether the target area exists in the maintained quadtree. In step 453, a node corresponding to the target area is placed. In step 455, the number of concurrent users of the placement node is increased by one. In step 461, it is determined whether the placement node requires division. If so, step 463 is executed. Otherwise, step 471 is executed. For example, if the updated number of concurrent users exceeds the region threshold, the child node of the placement node must be split. In step 463, four child nodes of the placement node are created and the number of concurrent users is obtained. In step 471, it is determined whether the placement node is a root node. If so, step 481 is executed. Otherwise, step 473 is executed. In step 473, the parent node is placed. In step 481, it is determined whether the slave server is overloaded. If so, step 483 is executed. Otherwise, the step is ended. For example, if the number of concurrent users of the root node exceeds the server load threshold, the maintenance area portion must be moved to another slave server or backup server. In step 483, the repartition request having the root node is transmitted to the VE management module 211. In step 485, the repartition result is received from the VE management module 211 and allows a transfer of control between the slave server and the backup server (if necessary) in the portion of the local area maintained. In step 491, a quadtree request is received. In step 493, the maintained quadtree is transmitted to the VE management module 211.

  In step 511, a root node repartition request is received. In step 513, the area corresponding to the received root node is defined as a congested area. In step 515, the quadtree request is transmitted to the corresponding slave server in the congested area. In step 517, the quadtree is received from the corresponding slave server. Note that each node of the received quadtree is marked as identical to the corresponding slave server, indicating that the region is processed by the slave server. In step 521, a quiet area is defined that is close to the congested area and has a minimum number of simultaneous users between the adjacent areas. This definition is achieved by detecting the number of concurrent users of the root node. In step 523, the number of simultaneous users in the quiet area and the congested area obtains the first number and the second number, respectively. In step 531, a root node corresponding to the congested area is arranged. In step 533, the next node corresponding to the current node is placed. The determination of the next node uses a horizontal traverse and a vertical traverse. In step 535, it is determined whether the placement node is a leaf node. If so, step 541 is executed; otherwise, step 533 is executed. In step 541, it is determined whether the area corresponding to the placement node is close to the quiet area. If so, step 543 is executed, otherwise, step 533 is executed. Neighboring nodes are preferably obtained using an effective quadtree traversal algorithm introduced by Samet. In step 543, the placement node is marked for merging with a quiet area. In step 545, the number of concurrent users of the placement node is added to the first number and subtracted from the second number. In step 551, it is determined whether the re-segmentation step is sufficient. If so, step 553 is executed, and if not, step 561 is executed. For example, if the second number is lower than the server load threshold, or a percentage of the server load threshold, the repartitioning process is sufficient. In step 553, the resulting quadtree is transmitted to the corresponding slave server.

  In step 561, it is determined whether the quiet area is full. If so, step 571 is executed. Otherwise, step 533 is performed. For example, if the first number exceeds the server load threshold or the percentage of the server load threshold, the quiet area is full. In step 571, it is determined whether there is a backup server. In that case, step 573 is executed. Otherwise, the process ends. In step 573, the next node corresponding to the current node is placed. It should be noted that the traverse approach utilized is similar to step 533. In step 575, it is determined whether the placement node is a leaf node. In that case, step 577 is executed, otherwise, step 573 is executed. In step 577, it is determined whether the area corresponding to the placement node is close to the quiet area. If so, step 578 is executed, and if not, step 573 is executed. In step 578, the placement node is marked with a backup server. In step 579, the number of concurrent users of the placement node is subtracted from the second number. In step 581, it is determined whether the re-segmentation step is sufficient, and if so, step 583 is executed. Otherwise, step 573 is executed. In step 583, the resulting quadtree is transmitted to the corresponding slave server and backup server.

  Details of the VE management method are explained below. 6a-6c are diagrams illustrating different exemplary quadtrees of the local region R6 maintained by the local region management module 221. FIG. In one example, referring to FIG. 6a, the local region R6 is necessarily divided into four parts R61, R62, R63, R64, and the subregion R63 is necessarily divided into four parts R631, R632, R633, And R634. The number of simultaneous users of R6, R61, R62, R63, R64, R631, R632, R633, and R634 is 30, 5, 5, 10, 10, 5, 2, 1, and 2. In step 411, a movement message indicating that the user U1 from the area R634 has moved to the area R64 is received by the local area management module 221. In steps 421, 423, and 425, the number of simultaneous users of the node N634 is decreased to one by one. In Steps 433 and 435, the number of simultaneous users of the node N63 is decreased to 9 one by one. The region threshold is set to 10, and in step 441, nodes N631, N632, N633, and N634 are determined to merge. In step 443, nodes N631, N632, N633, and N634 are deleted. In steps 431, 433, and 435, the number of simultaneous users of the root node N6 is decreased to 29 one by one. A temporary result is shown in FIG. Thereafter, in steps 431, 451, 453, and 455, the number of simultaneous users of the node R64 increases to 11 one by one. In step 461, it is determined whether node R64 is split. In step 463, four child nodes N641, N642, N644, N643 are created, and the number of simultaneous users is 4, 2, 2, 3. In steps 471, 473, and 455, the number of simultaneous users of the root node N6 increases to 30 one by one. Finally, the process is finished and the final result is shown in FIG.

  FIGS. 7a and 7b are diagrams illustrating different exemplary quadtrees of the local region R6 maintained by the local region management module 221. FIG. In another embodiment, referring to FIG. 7a, the local region R6 is necessarily divided into four parts R61, R62, R63, R64, and the subregion R64 is necessarily divided into four parts R641, Divided into R642, R643, and R644. The number of simultaneous users of R6, R61, R62, R63, R64, R641, R642, R643, and R644 is 30, 5, 5, 9, 11, 4, 2, 3, and 2. In step 411, a movement message indicating that the user U2 from another slave server has moved to the area R642 is received by the local area management module 221. In steps 421, 453, and 455, the number of simultaneous users of the node N642 is increased to 3 by one. In steps 461, 471, 473, and 455, the number of simultaneous users of the node N64 is increased to 12 one by one. In steps 461, 471, 473, and 455, the number of simultaneous users of the root node N6 increases to 31 one by one. The server load threshold is set to 30, and in step 481, the slave server 221 determines to overload. In step 483, the re-segmentation request for the root node N 6 is transmitted to the master server 21. In step 511, the re-partition request for the root node N 6 is received by the VE management module 211. In step 513, region R6 is defined as a congested region. In step 515, the quadtree request is transmitted to the local area management modules 221, 231, 241, 251 respectively. FIG. 8 is a diagram showing a typical quad tree maintained by the local area management modules 221, 231, 241, and 251. In step 517, quadtrees T6-T9 are received from local area management modules 221, 231, 241, 251. In step 521, region R7 is defined as a quiet region. In step 523, the first number 22 and the second number 31 are obtained. A vertical traverse is used, and nodes N61 and N62 are placed in steps 531 to 541. In step 543, the node N62 (that is, the region R62) is marked in the local region management module 231. In step 545, the first and second numbers are updated to 26 and 27, respectively. In step 551, it is determined that the re-segmentation step is sufficient. Finally, in step 553, the resulting quadtree is transmitted to the local area management modules 221 and 231 to switch the control of the area R62.

  In another example, at step 511, a repartition request having a root node N6 is received by the VE management module 211. In step 513, region R6 is defined as a congested region. In step 515, the quadtree request is transmitted to local area management modules 221, 231, 241, and 251 respectively. FIG. 9 is a diagram showing a typical quad tree maintained by the local area management modules 221, 231, 241, and 251. In step 517, quadtrees T6-T9 are received from local area management modules 221, 231, 241, 251. In step 521, region R7 is defined as a quiet region. In step 523, the first number 22 and the second number 38 are obtained. A vertical traverse is used, and nodes N61 and N62 are placed in steps 531 to 541. In step 543, the node N62 (that is, the region R62) is marked in the local region management module 231. In step 545, the first and second numbers are updated to 27 and 33, respectively. In step 551, it is determined that the re-segmentation step is insufficient. In step 561, it is determined that the quiet area is full because the first number exceeds 80% of the server load threshold. In steps 571-577, nodes N63, N64, N641, N642 are placed. In step 578, the node N642 is marked in the local area management module 261. In step 579, the second number is updated to 28. In step 581 it is determined that the re-segmentation step is sufficient. Finally, in step 583, the resulting quadtree is transmitted to the local region management modules 221, 231 and 261, and control of regions R62 and R642 can be switched. In the specific example described above, the received quad tree is processed like four independent trees of the master server 21, but the present invention is not limited to this. Those skilled in the art can merge the received quadtrees into a single quadtree and make appropriate modifications without departing from the scope and spirit of the present invention.

  DVE management systems and methods take the form of program code that is integrated into a tangible medium, such as a floppy disk, CD-ROM, hard drive, or other machine-readable storage medium (ie, Instructions), program code is loaded and executed by a machine such as a computer, which becomes a device for carrying out the invention. The method and system are integrated in the form of program code transmitted over a transmission medium, such as in the form of electrical wiring, cabling, optical fiber, or other transmission devices, and the program code is received, loaded, and computerized The machine becomes a device for carrying out the invention. When implemented on a general-purpose processor, the program code combines the processors to provide a unique device and, similarly, functions on specific logic circuits.

  In the present invention, preferred embodiments have been disclosed as described above. However, the present invention is not limited to the present invention, and any person who is familiar with the technology can use various methods within the spirit and scope of the present invention. Therefore, the protection scope of the present invention is based on the content specified in the claims.

It is a figure which shows the specific example of a DVE system. It is a figure which shows the hardware environment applied to a master, backup, and a slave server. It is a figure which shows the software mechanism by the example of a DVE management system. FIG. 3 illustrates a typical quad tree in a local area. It is a flowchart which shows the method of DVE management. It is a flowchart which shows the method of DVE management. It is a flowchart which shows the method of DVE management. It is a flowchart which shows the method of DVE management. It is a figure which shows the typical quadtree from which a local area | region differs. It is a figure which shows the typical quadtree from which a local area | region differs. It is a figure which shows the typical quadtree from which a local area | region differs. It is a figure which shows the typical quadtree from which a local area | region differs. It is a figure which shows the typical quadtree from which a local area | region differs. FIG. 3 illustrates a typical quad tree. FIG. 3 illustrates a typical quad tree. FIG. 3 illustrates a typical quad tree. FIG. 3 illustrates a typical quad tree.

Explanation of symbols

11 Processing Unit 12 Memory 13 Storage Device 14 Input Device 15 Output Device 16 Communication Device 17 Bus 21 Master Server 22, 23, 24, 25 Slave Server 26 Backup Server 31, 32, 33, 34 Four Areas 211 VE Management Module 221 261 Local area management module

Claims (10)

DVE management system,
It consists of a master server, receives a redistribution request and a first tree from the first slave server, a second tree from the second slave server, and the first tree consists of a plurality of first nodes, each of which It consists of the first number of concurrent users indicating whether the user is in the first area or a part of the first area, and the second tree is composed of a plurality of second nodes, each of which is the number of users. It consists of the number of second concurrent users indicating whether it is in two areas or part of the second area,
The first portion of the first area is close to the second area, and the number of second simultaneous users in the second area plus the number of first simultaneous users in the first area is the server load threshold. Or, if the ratio is lower than the server load threshold ratio, it is determined that the first part of the first area is controlled by the second slave server, and the part of the first area is controlled by the second slave server. And transmitting a result indicating that the control of the first portion of the first area between the first slave server and the second slave server is switched. The master server receives a third tree, and the third tree comprises a plurality of third nodes, each indicating a number of users in the third area or a part of the third area. It consists of three simultaneous users,
When the first portion of the first area is close to the third area, and the second number of concurrent users in the second area is less than the third number of concurrent users in the third area, the first area The system according to claim 1, wherein it is determined that the part is controlled by the second slave server.   The system of claim 1, wherein the first tree and the second tree are compatible with a quad tree.   The second part of the first area is close to the second area, the second number of simultaneous users of the second area, the first number of simultaneous users of the first part of the first area, and the first If the sum of the first concurrent users in the second part of the region exceeds the server load threshold or the server load threshold, the master server may The system according to claim 1, wherein the system is determined to be controlled by a server, and control of the first portion of the first area between the first slave server and the second slave server is changed.   The system according to claim 1, wherein the determination of the first portion of the first region is achieved by considering a horizontal traverse or a vertical traverse.   The first slave server transmits the repartition request when it is detected that the number of first concurrent users of the root node of the first tree exceeds a server load threshold. The system of claim 1.   When it is detected that the number of first concurrent users is lower than the region threshold, the first slave server deletes all child nodes for one of the first nodes, and the first concurrent users The system of claim 1, wherein four child nodes are created for region segmentation when it is detected that a number exceeds the region threshold.   When it is detected that the second number of concurrent users is lower than the region threshold, the second slave server deletes all child nodes for one of the second nodes, and the second concurrent users 8. The system of claim 7, wherein four child nodes are created for one of the second nodes when it is detected that a number exceeds the region threshold. Loaded and executed by the master server,
The master server receiving a repartition request from a first slave server;
The master server comprises a plurality of first nodes, and a first tree comprising a first number of simultaneous users indicating how many users are in the first area or a part of the first area, respectively. Receiving from one slave server;
The master server is composed of a plurality of second nodes, and a second tree composed of a second number of simultaneous users indicating how many users are in the second first area or a part of the second first area, respectively. Receiving from a second slave server;
The first part of the first area is close to the second area, and the first simultaneous user number of the first part of the first area is added to the second simultaneous user number of the second area. The master server determines that the first portion of the first region is controlled by the second slave server if the server load threshold or the server load threshold ratio is lower than
The master server transmits a result indicating that the portion of the first region is controlled by the second slave server;
Changing control of the first portion of the first region between the first slave server and the second slave server;
A DVE management method characterized by comprising: Loaded and executed by the master server,
Receiving a re-segmentation request from the first slave server;
A first tree consisting of a plurality of first nodes, each including a first number of simultaneous users indicating how many users are in the first area or part of the first area, is received from the first slave server. And the stage of
A second tree consisting of a plurality of second nodes, each consisting of a second number of simultaneous users indicating how many users are in the second first area or part of the second first area, from the second slave server Receiving, and
The first part of the first area is close to the second area, and the first simultaneous user number of the first part of the first area is added to the second simultaneous user number of the second area. Determining that the first portion of the first region is controlled by the second slave server if the server load threshold or the ratio of the server load threshold is lower,
Transmitting a result indicating that the portion of the first region is controlled by the second slave server;
Changing control of the first portion of the first region between the first slave server and the second slave server;
A DVE management computer code setting method characterized by comprising:

Images